Binaural panning overview
An important part of mixing audio signals is the placement of individual sound sources at different spatial positions. Most common recording and mixing techniques simply send a given signal at different levels to the available loudspeakers—two for stereo, or more for quadraphonic or surround setups—to create a virtual sound stage.
This approach is somewhat flawed, however, as human beings are able to locate sound sources at different positions with just two ears. Virtually all spatial information for all sounds is included in the two signals arriving at the two eardrums. From these signals, human beings can determine characteristics such as inter-aural time, level differences, and—based on the listening experience—information about the spatial origin of the sounds being heard. Are they coming from in front or behind, from the left or right, from above or below? This ability to perceive where a sound originates from is referred to as binaural hearing.
In theory, the spatial positioning of any sonic experience can be reproduced during playback, so no special techniques need to be employed during recording. There is, however, one drawback to this approach: every person has differently shaped ears and different body and head proportions, all of which influence the way sound signals arrive at the eardrum—not to mention aspects such as hearing loss, subjective responses to the sounds being heard, and so on. Given these physical differences, each person listening to the same sound source, while standing or sitting in the same position, will hear slightly different binaural signals.
Therefore, perfect reproduction would only be possible if you could make a recording with tiny microphones placed inside your auditory canals. Because this is impractical, binaural hearing has been emulated in sound laboratories by using mannequin heads with built-in microphones. This approach has led to binaural recordings to fit the average person, which are more or less compatible with the way most people hear.
Playback of binaural recordings is best suited to headphones, ideally combined with signal conditioning (processing) that ensures the signals are accurately reproduced. Given a specialized listening environment, it’s also possible to reproduce these signals with loudspeakers, utilizing a process known as crosstalk cancellation.
As you are unlikely to have the technology required to make binaural recordings, your best chance of simulating binaural signals is by processing the sound signal on playback. This is known as the head-related transfer function (HRTF), which approximates the change that a signal undergoes on its way from the source to the eardrum.